1

Flammability Investigation of Different Refrigerants using an operating MAC system

in a simulated front end collision situation

Martin Graz, Uwe Wuitz

OBRIST Engineering

Introduction

A flammability investigation of three different refrigerants for Mobile Air Conditioning was carried out. In a

first step the actual safety level of R134a was investigated. The results were put in relation to the two possible

alternatives R744 and 2,3,3,3-Tetrafluoroprop-1-ene (also called HFO-1234yf).

Test Setup – First Phase

A VW Lupo system was used to do the testing (Figures 1-2). The system was charged with refrigerant (500gr

for R134a and HFO-1234yf) and operated under real operating conditions (Pd ~15bar). As lubrication

medium PAG oil ND8 (135ml for all tests) was used. A possible leak of the refrigerant caused by a front end

collision was simulated by a release of refrigerant through a manually operated valve onto a hot surface. Such

a front end collision can rupture a refrigerant line and release the refrigerant-oil mixture into the engine

compartment. The released refrigerant-oil mixture was directed to a hot surface simulating a turbo charger or

hot exhaust manifold. The surface temperature was measured and the temperature was adjusted through a

controller of the electrical powered heat source (Figures 3-4).

Figure 1 R134a A/C system (VW Lupo) Figure 2 Test adaptations to R134a A/C system (VW Lupo)

The base line test with R134a (ignition temperature >743degC) and ND8 (flash point 204degC) was carried

out to define the existing safety level in the vehicle. A surface temperature of 970degC was chosen. Under this

condition it was possible to show that the mixture could not be ignited.

2

Figure 3 Test Procedure Figure 4 Test Background

R134a prevents the mist of oil and refrigerant vapor from ignition. During this test no flame propagation was

observed. It could be concluded that the use of R134a and ND8 oil is safe up to surface temperatures of

970degC.

In a second test, R134a was replaced by R744 (charge of 50gr and additional R744 container of 20kg was

connected to the system) and the test was repeated at the same surface temperature of 970degC. Since the oil

circulation rate in a R134a system is higher (~4%) and more refrigerant is solved in the circulating oil than in

an R744 system, the risk of mist ignition and the causing of a pilot fire is reduced with the inflammable

refrigerant R744. In a normal operating, R744 system pressures are higher than in the one used during the test.

However, the lower-than-normal pressures of the R744 system are considered to increase the risk of ignition

since a higher pressure would cause also higher release speeds and therefore reduce the probability to ignite

the mixture. In the test the release pressure was stabilized to 20bar. No misting was observed and no ignition

was observed.

The third test was carried out with 2,3,3,3-Tetrafluroprop-1-ene (also called HFO-1234yf and having a flash

point of ~400degC). Miscibility and solubility characteristics with ND8 are considered to be similar to R134a.

During the testing an ignition was observed and the flame propagation and the flammability envelop was

judged to be substantial.

Conclusion First Phase

R134, having an A1 rating (not flammable according to ASHRAE safety standard), cannot be ignited during

an accident simulation (vehicle front end) and under realistic operating conditions including circulating oil.

R744, having also an A1 rating, behaves similar to R134a and the risk of ignition can be judged to be lower

than with R134a. R744 is therefore generating an improved safety level. The HFO-1234yf, with a probable

ASHRAE safety rating of A2 or an even lower safety level in terms of flammability and toxicity (toxicity

testing not completed) clearly increases the risk of a pilot fire after a front end collision.

3

Test Setup – Second Phase

In a second level testing the release of HFO-1234yf oil mist onto a surface with lower temperatures was

investigated. The temperature was reduced to 800degC in order to simulate an operating turbo charger hit by a

mixture of tetrafluoro-1-propene and PAG oil ND8. In this scenario the mixture was again ignited. The flame

propagation and the flammability envelope lead to the conclusion that this pilot fire had caused a secondary

fire in the engine compartments. It could be concluded that by using HFO-1234yf, today’s high safety level in

terms of flammability would be significantly reduced, causing an increased vehicle damage after moderate

front end collisions and putting human life at risk.

In a next step, the temperature was reduced to 600degC (simulating a hot exhaust manifold). Also in this setup

the refrigerant oil mixture was ignited. The flame propagation was substantial and even so the flammability

envelope was reduced, it could still be judged to be sufficient to cause a secondary fire.

Conclusion Second Phase

It can be concluded that 2,3,3,3-Tetrafluoroprop-1-ene (HFO-1234yf) used as a refrigerant in a vehicle

application reduces significantly today’s safety level in terms of flammability. Possible implications are

increased insurance premiums since the cost for repair will be increased. Further investigation is necessary on

the implications for persons being present in confined spaces, e.g. during accidents in tunnel or garages, when

such a fire happens (due to the low LC50 values of the thermal decomposition products). The risk of a pilot

fire caused by a front end collision in a vehicle using HFO-1234yf is significantly increased. A risk that the

pilot fire causes a secondary fire became clearly visible during testing and is highly possible (the flames

consumed the polycarbonate indicator panels and the operational wire harness).

Background

Pure ND8 oil was released without refrigerant to the 600degC surface. The oil was ignited. The flame

propagation and the flammability envelope were judged to be moderate but the burning oil will provide the

MIE (minimum ignition energy) for the ignition of HFO-1234yf.

An additional flammability test of pure HFO-1234yf was conducted using a cigarette lighter. The refrigerant

was ignited and flame propagation was observed. The person handling the test was equipped with an active

carbon filter mask and fire protective gear. The reason for the mask and protective gear was that the thermal

decomposition products of HFO-1234yf are Hydrogen fluoride (HF) and Carbonyl fluoride (COF2). These

substances have a LC50 (lethal concentration) value at 1h of 966 ppm and Carbonyl fluoride (COF2) is

4

having a LC50 value at 1h of only 360 ppm. LC50: 50% of an animal population dies within 1h if the

atmosphere contains the indicated amount of the substance in ppm.

Technical Development Direction

When compared to R152a (A2), the HFO-1234yf flammability envelope is slightly reduced. However, the

flammability envelope of HFO-1234yf requires either an oil free circuit or a secondary loop for the front end

heat exchanger.

Outlook

In a next phase, side impact collision will be investigated during which refrigerant lines inside the passenger

compartment are ruptured.

Annex

1) Material certificate for 2,3,3,3-Tetrafluoroprop-1-ene CAS No 754-12-1

2) MSDS for 2,3,3,3-Tetrafluoroprop-1-ene Chemical Formula C3H2F4

3) MSDS for PAG oil ND8

4) MSDS for R744

5) MSDS for R134a

6) 2,3,3,3-Tetrafluoroprop-1-ene Container Sticker pictures

5

1) Material certificate for 2,3,3,3-Tetrafluoroprop-1-ene CAS No 754-12-1

Product: 2,3,3,3-Tetrafluoroprop-1-ene

Code No: PC0987

CAS No: 754-12-1

Formula: C3H2F4

Batch No: Q14B-40

Purity: 99%

Identity: Conforms to reference

Date: 08th February 2008

CERTIFICATE OF ANALYSIS

6

2) MSDS for 2,3,3,3-Tetrafluoroprop-1-ene Chemical Formula C3H2F4

7

8

9

10

3) MSDS for PAG oil ND8

11

12

13

14

15

4) MSDS for R744

16

17

18

19

5) MSDS for R134a

20

21

22

23

24

25

26

27

28

6) 2,3,3,3-Tetrafluoroprop-1-ene container sticker picture